Self-correcting vertical flatness in a glass tempering system

Patent metadata

Field	Value
Publication number	US-12560920-B2
Application number	US-202217961717-A
Country	US
Kind code	B2
Filing date	Oct 7, 2022
Priority date	Jan 6, 2022
Publication date	Feb 24, 2026
Grant date	Feb 24, 2026

Abstract

This disclosure is directed to techniques for utilizing various sensors and models to evaluate glass as it progresses through the tempering process in order to ensure that the tempered glass is of a proper quality. If, according to any of the various sensor measurements, the tempered glass is not of a proper quality, the system may automatically adjust one or more settings in any of the various components of the system in order to bring future panes of tempered glass back to having the proper quality. The system can measure for any number of glass characteristics or system characteristics, including edge quality, vertical flatness, haze, washing process variables, thermal imaging, distortion, blower information, production data, and furnace process data.

First claim

What is claimed is:

1 . A method comprising:
after a first sheet of tempered glass exits a furnace system operating under an initial set of one or more furnace parameters, and while the first sheet of tempered glass is in a vertical position, controlling, by one or more processors, a scanner system to measure one or more vertical flatness data points for the first sheet of tempered glass; applying, by the one or more processors, a machine learning model to determine whether to make one or more adjustments to the initial set of one or more furnace parameters based on the one or more vertical flatness data points for the first sheet of tempered glass, wherein the machine learning model comprises historical data for the furnace system; in response to determining to make the one or more adjustments to the initial set of one or more furnace parameters, automatically adjusting, by the one or more processors, one or more furnace parameters in the initial set of one or more furnace parameters to develop an updated set of one or more furnace parameters; and after applying the machine learning model, updating, by the one or more processors, the machine learning model to develop an updated machine learning model, wherein the updated machine learning model includes an entry comprising at least the initial set of one or more furnace parameters and the one or more vertical flatness data points.

2 . The method of claim 1 , further comprising:
after a second sheet of tempered glass exits the furnace system operating under the updated set of one or more furnace parameters, and while the second sheet of tempered glass is in the vertical position, controlling, by the one or more processors, the scanner system to measure one or more vertical flatness data points for the second sheet of tempered glass; applying, by the one or more processors, the updated machine learning model to determine whether to make one or more adjustments to the updated set of furnace parameters based on the one or more vertical flatness data points for the second sheet of tempered glass; and in response to determining to make the one or more adjustments to the updated set of one or more furnace parameters, automatically adjusting, by the one or more processors, one or more furnace parameters in the updated set of one more furnace parameters to develop a second updated set of one or more furnace parameters.

3 . The method of claim 1 , wherein automatically adjusting the one or more furnace parameters in the initial set of one or more furnace parameters comprises controlling, by the one or more processors, the furnace system such that the furnace system begins operating under the updated set of one or more furnace parameters.

4 . The method of claim 1 , wherein the one or more vertical flatness data points include one or more of a load pattern for the first sheet of tempered glass, a glass type for the first sheet of tempered glass, and a size of the first sheet of tempered glass.

5 . The method of claim 4 , wherein applying the machine learning model comprises:
comparing, by the one or more processors, the load pattern for the first sheet of tempered glass to one or more historical load patterns for sheets of tempered glass with a same glass type and a same size as the first sheet of tempered glass; comparing, by the one or more processors, the initial set of one or more furnace parameters to historical furnace parameters for the sheets of tempered glass with the same glass type and the same size as the first sheet of tempered glass; and based on the load pattern comparison and the furnace parameter comparison, determining, by the one or more processors, whether to adjust the initial set of furnace parameters.

6 . The method of claim 1 , further comprising:
after a second sheet of tempered glass exits the furnace system operating under the updated set of one or more furnace parameters, and while the second sheet of tempered glass is in the vertical position, controlling, by the one or more processors, the scanner system to measure one or more vertical flatness data points for the second sheet of tempered glass; applying, by the one or more processors, the machine learning model to determine whether to make one or more adjustments to the updated set of one or more furnace parameters based on the one or more vertical flatness data points for the second sheet of tempered glass; and in response to determining to not make the one or more adjustments to the updated set of one or more furnace parameters, controlling, by the one or more processors, the furnace system to continue operating according to the updated set of one or more furnace parameters.

7 . The method of claim 1 , wherein the furnace system comprises one or more of a heating component and a quench component.

8 . The method of claim 7 , wherein the heating component comprises a tempering oven.

9 . The method of claim 7 , wherein the quench component comprises a series of one or more air blowers.

10 . The method of claim 1 , wherein the initial set of one or more furnace parameters comprises one or more of heating temperature, loading delay, heating time, convection values, air balance, nozzle distance, quench pressure, roller speed, cooling temperature, and fan speed.

11 . A device comprising:
a memory; and one or more processors configured to:
after a first sheet of tempered glass exits a furnace system operating under an initial set of one or more furnace parameters, and while the first sheet of tempered glass is in a vertical position, control a scanner system to measure one or more vertical flatness data points for the first sheet of tempered glass;
apply a machine learning model to determine whether to make one or more adjustments to the initial set of one or more furnace parameters based on the one or more vertical flatness data points for the first sheet of tempered glass;
in response to determining to make the one or more adjustments to the initial set of one or more furnace parameters, automatically adjust one or more furnace parameters in the initial set of one or more furnace parameters to develop an updated set of one or more furnace parameters;
after a second sheet of tempered glass exits the furnace system operating under the updated set of one or more furnace parameters, and while the second sheet of tempered glass is in the vertical position, control the scanner system to measure one or more vertical flatness data points for the second sheet of tempered glass;
apply the machine learning model to determine whether to make one or more adjustments to the updated set of one or more furnace parameters based on the one or more vertical flatness data points for the second sheet of tempered glass; and
in response to determining to not make the one or more adjustments to the updated set of one or more furnace parameters, control the furnace system to continue operating according to the updated set of one or more furnace parameters.

12 . The device of claim 11 , wherein the machine learning model comprises historical data for the furnace system, and wherein the one or more processors are further configured to:
after applying the machine learning model, update the machine learning model to develop an updated machine learning model, wherein the updated machine learning model includes an entry comprising at least the initial set of one or more furnace parameters and the one or more vertical flatness data points; after a second sheet of tempered glass exits the furnace system operating under the updated set of one or more furnace parameters, and while the second sheet of tempered glass is in the vertical position, control th

Assignees

• Cardinal Ig Co

Inventors

• Pylkkönen Jarmo T
• Gursky Jacob N

Classifications

• G05B19/4183

  characterised by data acquisition, e.g. workpiece identification · CPC title

• G06T7/0004

  Industrial image inspection · CPC title

• C03B29/08

  Glass sheets · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title

• G01N21/8901

  Optical details; Scanning details (per se G02B) · CPC title